Following the discussion in this thread, I decided to dig in and do some research on the exact wear characteristics and likely lifespan of FRC batteries. I compiled my findings into a document here.
I wont repost the entire thing here, but here’s my takeaways:
Lead Acid batteries suck, lasting somewhere between 20 and 100 cycles before losing enough performance to be considered unhealthy.
Desulfating batteries can significantly increase lifespan into the upper end of that range.
Increasing charge current to 1C and beyond can greatly increase lifespan.
Decreasing Depth of Discharge can increase cycle life and overall throughput.
Using ESR as a metric for battery health is acceptable if desulfated.
If you have questions, take a look at the full writeup first to see if my logic makes sense. There’s definitely a little inconsistent information even between recent papers, so I had to make some judgement calls on things like trusting academics vs. companies. There are many confounding factors that go into evaluating SLA batteries. Overall though, I feel happy with my conclusions, and hopefully this writeup is useful to teams looking to make the most of their batteries.
As an aside, I’ve been doing plenty of research into LiFePO4s, and hope to have more information on them soon. My new job gives me access to some really nice battery cell testers.
One thing I want to know more about specifically is the relationship between voltage and state of charge. There is a nice linear relationship over part of the life of an FRC battery but I’m not sure it holds completely and I think there is a more realistic model that can be derived from current used and the voltage.
The manual for the DualPro RS3 chargers from AndyMark doesn’t go into much detail on how they work. It’s 6A per bank (if I understand correctly this would be C/3 for our batteries?) and there’s a brief mention about sulfation suggesting you should leave the batteries plugged in “whenever the opportunity arises.”
One popular brand of charger is NOCO Genius, which offer 1, 2, 5, 10, and 25-amp chargers, with the latter being a lot of money. The 10-amp 4-bank charger is a more reasonable price than the 25-amp 1-bank charger. All of these advertise that they’re desulfating chargers.
I’m curious if you have a recommendation for charger specs. The charge rate you’re talking about (1C and beyond) seems prohibitively expensive.
New comp meta if this proves to drastically alter batt performance. Everyone charging batteries at hotels and running back and forth or finding other “creative ways” to charge the batteries
Maybe this was what the truck ad during Kickoff was for - an electric truck you can effectively use as a power bank could be useful if you want to skirt around the charging rules.
If you just want a first-order estimate, using a look up table with voltage at low loads will make it easy to find the SoC. I recommend looking at the OSU paper, as it describes both a simple second order battery model as well as the method to find parameters.
Characterizing this is a little bit of a challenge. Low like logic circuits only (idle robot) or low with actual measurement?
I started thinking about trying to put a current transducer on the main 12 volt line, and I’ll be surprised if we can find a DC sensor that both captures logic level draw accurately and won’t saturate at >150A Sprint current draws.
An inline resistor feels like it would be out of the question for performance reasons, along with getting the heat sink sufficiently sized.
I haven’t checked if the PDH has a global current draw API, I don’t remember one on the PDP.
Actually, the way I read R604, which is a robot construction rule, is that you can’t put a battery in the robot (on the field at a competition) if it’s been charged above 6A. Obviously it’s a bit subjective, though.
The lowest loads you can find, if possible. For in-match use, you’ll need to use the OSU paper model. SLA batteries have large time constants, so you’d have to sit for 10 seconds to get a good reading.
Agreed, no economical sensor exists that can accurately measure .1A to 400A (A hole shot with 4 Falcons can get to 400A for 10s of milliseconds, maybe a bit longer) however, you can measure the quiescent (idle) current with a basic DVM. Write it down. Then use something like this to big haus to do the heavy lifting. https://www.digikey.com/en/products/detail/lem-usa-inc/HAS-400-S/1026544
The quiescent current is a rounding error for the 2 minutes of a match and really only matters if the robot is left on for a long time.
You could also add a more sensitive sensor on the line along with the big haus and read it when the current is within its range and let it saturate during the big hits. Not sure if extreme saturation would damage such a sensor.
What part of R604 is subjective in your view? Seems rather cut and dry (in my reading of it) that if you’re charging your batteries at a rate greater than 6A at the competition, you’re in violation of said rule.
I do think it’s odd that R604 is listed in the Robot construction Rules rather than event rules.
Note that the green text uses the word “safe” which implies an intent of safety rather than fairness.
I think the subjectivity stems from the placement of the rule in the robot construction rules rather than the event rules, as you said.
“ROBOT battery” can clearly only refer to the battery on the robot at the competition.
Let me put it this way… if a team had a 25A charger at the competition and used it to charge the battery powering the LEDs on their robot cart (which is a standard 18 Ah battery from a couple seasons ago) then this rule wouldn’t apply because it’s not the battery going on the robot.
But the implication (by using the word “safe” in the green text) is that they don’t want you charging these FRC batteries at more then 6A at the competition.
If that’s their intent, I suggest moving the rule to the event rules.
I’m curious as to the reasoning HQ has chosen to put this rule with the robot construction rules… I doubt we’ll get an answer on these forums, but still would love to know.
When you stop applying power to a motor and there is momentum still moving the motor, the motor becomes a generator and sends power back to the battery. In EVs this gets called regen and is probably a bit fancier both software and harware wise than in FRC in order to be more efficient. I should probably run some calculations to make sure but I’m guessing it would be trivial to exceed 6A battery charging current, even if only for a moment, with an FRC robot during a match via regen.
Still, I am joking and not seriously impling that this really is or should be a violation of R604.
Don’t know why the founding FRC fathers made this rule, but if they removed it we would still use the 6A max (our current chargers are 4A) since at competition, charging 12 batteries @ 6 Amps plus any other pit power starts to tax the 15A breaker powering the pits.
I think they wanted to keep car battery chargers that have the quick boost charge out of the event. If left unattended and on boost bad things can happen. One of the guys at work did it to a similar sized AGM emergency light battery and when he remembered 2 hours later the battery was round, actually didn’t split apart but it inflated like a beach ball
I doubt we’ll get an answer on these forums, but still would love to know.